Although early meteorological records were naturally confined to the weather experienced at the surface of the Earth, as soon as meteorology began to be a science it was realized that what went on above the surface was important and that the physics of the air needed to be studied in three dimensions. An obvious illustration is the use made of cloud observations—the nature and structure of clouds, the method of their formation and their movement. Something could be learnt on those questions by observations from the ground, aided by such a simple instrument as the nephoscope. Most people, and certainly all concerned with navigation, know that the movement of clouds may be very different from the wind at the surface. It was also realized that since the atmosphere is a heat engine a knowledge of the temperature of the upper air was required. Anyone who has climbed a mountain knows that the temperature falls, and this was confirmed for the free atmosphere by observations with kites. It was clearly of interest to know whether the fall of temperature was maintained until the absolute zero was reached, and if not, why not. It is in the measurement of temperature and wind in the upper air that balloons have found their chief use.